I think what's doing a lot of the work in the Jane case is that intuitions are being pumped by the weirdness of your situation.
Unless you posit that everyone is being made by crazy doctors, your situation is a very weird one, even if there are lots of crazy doctors. Being made by a crazy doctor is not a precondition for an existence of someone like you. Unlike fine-tuning.
So there's kind of a reference class problem. If you take the most intrinsically significant thing about your identity to be that you were made by a crazy doctor, then you might want to posit lots of crazy doctors so that somebody like you can exist. But if you take this fact about your origin to be insignficant and incidental to your identity, then you'll be inclined to disregard it and see it as something that doesn't need an explanation.
In the actual fine-tuning case, what is salient or significant is that conscious observers and life exists at all. That does not seem to be so amenable to treatment as an insigificant feature of the universe that only matters to us because it happened to take place, though some opponents of the FTA want to say so. We look at the evidence for fine-tuning, and it strikes us that this is really important. That a universe without fine-tuning would be utterly different. That this is really important and demands an explanation.
That you were made by a crazy doctor? Not so much. It's weird, but it doesn't strike us as all that significant. It doesn't demand an explanation in the same way. It's more like being born in an aeroplane, or being conceived as a consequence of your parent going on a holiday they could only afford because they won the lottery. There are all sorts of stories about how people are conceived and born, and they're all unique. Nothing about Jane really cries out for an explanation in particular. Every conception is a miracle. Jane's is no different to anybody else's.
But if Jane doesn't know if there are other humans, and if Jane thinks that humans can only be created by crazy doctors, then she may have reason to think there are many crazy doctors. Or if Jane thinks of people who are made by crazy doctors as being importantly different from regular people, as different as we are from bacteria, then she might think that there are lots of crazy doctors.
Interesting! How would this figure in the Bayesian calculation? Different prior? If so, then you still ought to get an evidential boost for Many Doctors, but that doesn't seem right.
This is exactly right. Philip would object that the existence of other crazy doctors does not increase the probability of HER existence. But when we compare physical theories, the relevant evidence is class evidence, not token evidence. So the right perspective is the “eye-in-the-sky” one:
"Alan is a spy who wants to find out how many doctors work at the IVF institute. Alan, like everyone else, knows that every doctor in the country has a strange disorder: each doctor performs at most one IVF procedure, and only if they roll double six on the first try. Alan waits in the bushes near the institute for several years. He never sees the doctors, because they use a secret tunnel. The only information available to him is whether babies are brought out. One day, Alan observes that a baby is brought out—Jane. He reasons as follows: if there is only one doctor, the probability that a baby is born is 1/36; if there are many doctors, the probability is much higher (approaching 1 as the number of doctors increases). Hence, given that a baby is brought out, it is more likely that there are many doctors." (https://philpapers.org/rec/NABTMV)
It (life in the universe) does not “demand an explanation” at all. This is the fundamental statistical error of your entire argument. Giant statistical probabilities do not come with ‘explanations’. They are just what they are. You cannot use statistics in this way and this is what people just don't understand.
If you win the lottery, then, make sure you return the ticket because it was clearly statistically impossible to any reasonable measure that you just won, and that you demand a logical explanation for it before accepting your winnings.
Thank you for illustrating your point so I can respond.
When I say something "demands an explanation", I just mean you should seek an explanation. You're not guaranteed to find one.
"Statistically impossible" just means improbable. Winning the lottery is very very improbable. If you appear to win the lottery, you *should* be asking yourself if winning the lottery is really the best explanation for your evidence. You should be wondering if you're dreaming or the victim of a prank. You should, in other words, be looking for an explanation.
If one is not forthcoming, then eventually you just have to accept that the best explanation of the evidence is that you have actually won the lottery. But that's only because there is no better explanation. It is never the right attitude to just ascribe some very unlikely event to a freak coincidence when alternative explanations are more likely.
Your problem is that you're basing your beliefs upon this statistical idea you've discovered (the universes precise parameters). But like I'm trying to explain, these really big statistical measures literally do not work in the way you think they do. You can't even say you 'should' look for an explanation. There is very good modern analytical philosophy on statistics I could recommend you because many people, even philosophers make this mistake.
In every day life, if, say, the 80 year old president of a country suddenly does a speech in Polish, having shown no prior evidence of knowing our speaking Polish, then that is a unusual event that we should, and could, look for an explanation for. There would also be an explanation we could find. But this rare and odd event is *not* analogous to modern statistical measurements (of gigantic numbers), for evident reasons that can be explained. Statistics are not giving you the same type of information. The fine tuning argument is the same thing. It's not giving you the evidence for something rare that you think it is. It's purely statistical. You can't use this as actual evidence of anything.
I do not agree that there is a relevant difference between spotting something about a numerical distribution and spotting something odd about a president speaking a language.
What exactly is the difference you’re trying to point to? What criterion are you using? Is it just that numbers are involved? If I roll a die 10 times and get only sixes, is that evidence that the die is not fair?
I think that this is correct: what's doing at least some of the work here is that our background prior on "multitude of crazy doctors rolling dice" is so low compared to our prior on the "the doctor cheated/dice were rigged" hypothesis because rigged dice are relatively common.
But if Jane does some investigating and finds that these were, idk, quantum dice, such that rigging the dice would have required someone to modify the most fundamental forces of reality, then surely this does give Jane evidence for the many doctors hypothesis over the "cheating" hypothesis
My intuition is that it doesn't provide any boost at all. But if you don't share that intuition, we have to turn to the broader analysis of what's going on in different cases, that Kenny Boyce, Mark and I has been doing
When you say 'teleological universe', how would this fit, or not, with Lee Smolin's and Rupert Sheldrake's (as best I know independent) proposals that natural laws evolve? Does the fine tuning depend on this being a tunable instrument? And if so, do we find an example of self-tuning in homeostatic organisms?
I have an additional issue with the multiverse theory. As an argument against ‘something godish’ it has the same weakness as the argument that the laws of nature, or for that matter evolution, dispenses with the need for something godish.
Seventeenth century scientists developed the idea that the laws of nature are the regularities with which God governs the world. By the middle of the eighteenth century some people including David Hume could take for granted that the laws of nature are self-explanatory, and therefore debate whether there is a god at all. But Hume spotted the problem. As an empiricist he noted that we never observe causation. We infer it. The laws of nature are the regularities, not the causes of the regularities. Once this point is seen it can be applied to evolution and multiverses just as well. Any or all of the three could be God’s way of creating stuff. If they aren’t, we still have no account of causation, just a map of successive events.
This assumes a Humean view where laws are merely descriptive summaries of what happens. That is highly controversial. Frameworks like Constructor Theory and Modal Realism suggest instead that laws are grounded in underlying governing principles. If these principles are the inherent architecture of a concrete world, they are the cause of the regularities.
How does this help? If the laws of nature are the inherent architecture of a concrete world, how does calling them the inherent architecture make any difference? (And why should Hume disagree?) We can still ask what caused them to be there. In other words, how have you done more than just add another layer of definition?
It helps in the sense that it reduces physical laws to ontological, logical necessities. Modal Realism and Constructor Theory propose that the 'architecture'—the underlying laws of what is possible and impossible—is the fundamental floor. At some point, one hits the 'Brute Fact' of logical and physical consistency, which is a far more robust explanation for reality than Hume’s 'infinite coincidences' or the necessity of a creator.
So the ‘architecture’ is the ‘fundamental floor’. It relates to the physical laws by being different from them and being the cause of them. As you say this is a theory, not accessible to empirical evidence. Use of the term ‘floor’ suggests to me that the theory denies that it is in turn caused by something else. It sounds to me like an ‘uncaused cause’. If I have understood this correctly, I’d like to ask how it differs from Thomas Aquinas’ uncaused cause, namely God. You say it is ‘far more robust’ than either Hume or ‘the necessity of a creator’. How is it more robust? For Aquinas the uncaused cause was an agent, more intelligent than humans, deciding to create. To enquire further we must reflect on the nature of God, which people have done all through history. Your fundamental floor, as I understand it, is ‘just there’. To me this conjures up a picture of an implied noticeboard forbidding us to enquire any further. Isn’t that rather unscientific? Or is there more to it?
While Modal Realism is metaphysical, Constructor Theory provides a framework for potentially testable predictions. In both cases, the 'fundamental floor' is logical necessity. A purported deity either equates with logic or is subject to it. In the first case, this is merely ontological relabeling. In the second, it introduces an extra entity not required by the theory, and is therefore, by Occam’s razor, less robust.
Furthermore, I’m curious about your stance on the nature of the 'Uncaused Cause.' In Aquinas’s time, little was known about the mechanics of intelligence, but modern science (neuroscience, AI) reveals that intelligence is a product of immense functional complexity. If you hypothesize a 'highly intelligent' deity, do you regard them as a complex entity or a monadic, simple one?
If the former, we must ask what 'architecture' allows for that complexity; if the latter, you are positing a type of intelligence that contradicts everything we currently understand about information processing and requires a nomologically radical hypothesis—making it far less robust than nomologically conservative theories like Constructor Theory or Modal Realism.
Thank you. I find this fascinating and am happy to continue the dialogue, if you are willing. But please ignore me if I’ve passed my Best Before date.
So the ‘fundamental floor’ is logical necessity. Does this mean that what is logically impossible cannot exist? I think this exaggerates the power of human logic. Lots of theories have proved true even though they defied logic at the time: wave v particle theories of light, the Einstein-Podolski-Rosen experiments with electrons, etc. I’m sure you can think of others. At the other extreme there are theorists, like the evolutionary psychologist Patricia Churchland, arguing that the human mind has only evolved for the four ‘f’s (feeding, fighting, fleeing and reproducing) so ‘Truth, whatever that is, takes the hindmost’. (I’m not defending her. It’s just that your position sounds to me like the other extreme.) So when you say ‘A purported deity either equates with logic or is subject to it’ you seem to me to be denying that something can exist which transcends human understanding. For the purposes of empirical research based on the scientific method this is usually good enough – but not for all purposes.
On Ockham’s razor. It’s a useful heuristic tool, but only a tool. Everything from the molecules in Ockham’s fingernails to the Milky Way has turned out to be far more complex than anyone knew in his day. The simplest solution isn’t always the right one, to say the least. So when you speak of ‘extra’ entities ‘not required by the theory’, research doesn’t limit itself to the current theories, at least not in the long term.
On divine creators. They have been hypothesised since prehistoric times in every human society as far as anthropologists have observed. So when you describe them as a ‘nomologically radical hypothesis’ I wonder what you mean. Your ‘conservative theory’, as a replacement, only goes back to eighteenth century France. More important, people have believed in divine creators for other reasons too – but that would take us onto other topics.
I wonder whether the underlying difference between us is that where you respond to the prospect of a divine intelligence by asking what ‘fundamental floor’ could allow for it, I’m responding to the prospect of a ‘fundamental floor’ by asking what intelligence could allow for it. In a sense your foundation is stuff, mine is intelligence. Does that sound right to you?
I don't see that as a problem with the multiverse explanation. That's a separate problem. The problem of where the laws come from, or why anything exists. The multiverse is only posited to explain fine-tuning. It doesn't have to explain everything. If you want to take the existence of the laws as evidence for God, that's fair game. But it would probably invite different responses from the atheist side (which would likely be compatible with the multiverse explanation).
I don't think an appeal to the anthropic principle actually requires multiple universes?
The basic argument is that if the universe wasn't fine tuned for life, we wouldn't be here to discuss it, so we can't infer anything about design from the fact that the universe is fine-tuned. Whether life was extremely improbable (fluke occurance) or guaranteed in some universe (multiverse), the argument still works.
It would be like if Dr. Smith later told Jane this story, and didn't mention what the dice rolls were, just mentioned that if it hadn't been all sixes, Jane wouldn't have been born. Jane can infer from this that Dr. Smith must have rolled all sixes, even without asking.
Your earlier Scripture in the Big Bang argument is a lot more interesting in my opinion, since that would be a fact about the universe which has no obvious connection to the existence of intelligent life.
The real question is: Can we come up with any way the universe is "fine-tuned" that isn't contingent on the same things that intelligent life is contingent on? Possibly.
There are different versions of the anthropic principle, but as far as I can tell most people do not find the kind of explanation you are offering to be satisfactory.
If the universe wasn’t fine-tuned, we wouldn’t be here to discuss it. This means we know the universe must be fine-tuned. It doesn’t mean that this is unsurprising.
If I tell you that I am secretly an alter ego of Donald Trump, then (assuming you believe me), you know that I must be the president of the USA. It doesn’t mean that this is unsurprising. Why would Donald Trump be posting on substack about philosophy? It doesn’t seem like something he would do.
The research question: Are universes fine tuned for life?
The sample size is 1.
The sample is biased because it wouldn't exist unless its universe was fine tuned for life.
irrelevant hypothetical a-priori chance a universe is fine tuned for life: 1 in 100^100
The result: The universe is fine tuned.
Since the sample size is 1 and the result is 1 out of 1, the p-value would be 1 in 100^100. Except the problem: The sample is biased. Due to this bias, to be intellectually honest, we have to throw out the only data point. This leaves us without a p-value.
The multiverse or not multiverse question is a question of population size, not sample size, so it really has little statistical relevance here. Unless you are trying to be *extremely* precise, you don't need to know the population size to do statistics.
A biased sample does not force us to discard the data point. It forces us to model the selection effect properly, so that the evidence can still function as objective evidence.
The multiverse question is not about population size. It is about which hypothesis makes the existence of the relevant kind of observer-containing, fine-tuned universe more probable. Once the evidential setup is reconstructed in a modal, third-person way, the result is still confirmation of a multiverse over a single-universe hypothesis.
I don't think that is the research question. The research question is "Why is this universe fine-tuned for life?"
We don't know the population size. So we're debating whether we can explain it by a biased sample from a large population which is bound to have some universe fine-tuned for life, or as an unbiased sample.
You and I both think the sample is relevantly biased. But that is just the point of contention.
Philip and Kenny essentially think that this is the only universe we could have sampled, so the choice of universe is not biased. I think that because only fine-tuned universes can be sampled, then it is biased. Which way you look at it depends on whether you're more focused on who exactly is doing the sampling or on what kind of thing can be sampled.
Then Philip and Kenny assume for the sake of argument that we could have sampled other universes. But then their argument shows that fine-tuning has nothing to do with the inference to a multiverse. No matter how fragile or robust the conditions for life to be possible, we're more likely to be sampling something the more universes there are.
The question I am asking isn't whether a single universe or multiverse is more likely. You're right that the unlikelihood of fine tuning is a good argument for the multiverse as compared to a single fluke universe. I agree with you and don't contest that point. That is not my point.
What I am contesting is the idea that a single universe that happens to be fine tuned implies design. In my opinion, it doesn't imply design, assuming that the fine tuning can be shown to be a per-requisite for anyone asking this question in the first place.
If I understand you correctly, your view is that since we exist, the observational datum already guarantees that the universe is fine-tuned, so the probability of a fine-tuned single universe should be treated as 1. But that seems to conflate epistemic reasoning with ontological reasoning: from the fact that we now know we exist, it does not follow that a fine-tuned universe was not antecedently improbable. It only follows that our evidence is conditional on our existence.
That said, I am sympathetic to a modal-realist-like metaphysical view on which fine-tuning needs no further explanation, but that is beside the point here.
If the Qur'an were written out in the cosmic microwave background, that would imply design, even if that being written out in the cosmic microwave background would be a pre-requisite for anyone asking "Hey, how come the Qur'an is written out in the microwave background?".
I think it's actually even worse than that. If what we mean by "life-permitting" amounts to the physical parameters being such that stars can form and life can arise in a way that doesn't violate laws of physics, then on omni-theism, God *could* create life much like us via miraculous intervention in a non-life-permitting universe. He doesn't actually need a fine-tuned/life-permitting universe to create life like us. So, yes if naturalism is true and life exists, it's nearly a guarantee the universe is "life permitting." But if omni-theism is true, this is not a guarantee because we may be here via miraculous intervention (contra physical law).
Here's the question: Forget all this chemistry and biology giving rise to life.
Can some type of non-living consciousness exist in a chaotic and disordered universe?
If it can, the real mystery is why the universe is so ordered and intelligible (and if Einstein couldn't answer this question, then probably I can't either!)
The anthropic principle sortof explains why consciousness must exist, but it doesn't explain why reality isn't a never-ending acid trip.
For the fine-tuning argument, the relevant reference class is not just “conscious beings,” but conscious epistemic systems capable of discovering that they inhabit a fine-tuned universe and reasoning about it—otherwise we would not be having this discussion in the first place. In that sense, radically chaotic or unintelligible universes are not relevant counterexamples here, because they would not support the kind of epistemic system whose evidence is at issue.
Assuming some version of origin essentialism is true, God couldn't create us in a different universe. Our origins is necessarily tied to who we are. So if God wants to actualize us, He has to fine tune the universe. And this is also why we couldn't have existed in another universe.
I've never understood the premise that life is in special need of explanation any more than if the universe contained nothing. The anthropic principle strikes me as ironic in that, at the same time as it tries to make humans non-special (we're on one of many planets, in one of many universes), it begins with the anthropocentric assumption that humans need a special accounting.
But while the Many Doctors Hypothesis is clearly wrong, isn't it because Jane just shouldn't be asking the question about the improbability of her birth, which, retrospectively, was inevitable? Wouldn't the analogy to the fine-tuning argument for god suggest that Jane should think the dice were rigged?
I think it's enough for the coincidence to be subjective. The idea that this sort of coincidence has to be objective to warrant explaining is a mistake, in my view.
Otherwise, what you're saying suggests you would find nothing remarkable or in need of explanation in finding the Qur'an in the cosmic microwave background.
You could say that the fact that it is meaningful to us is just an artifact of the fact that we happen to have an encoding, a script and language on Earth (Unicode, Arabic script and language) that makes it seem meaningful. To an alien race with no such script or language the markings would look like random noise.
So there is nothing objectively remarkable or in need of explanation about finding the Qur'an clearly spelled out in the CMB.
If that's what you think, then I guess we'll just have to part ways, because this is clearly nuts to me. It's a wilful disregarding of very clear evidence.
On the other hand, if it is fair game to find it remarkable that the Qur'an is clearly in the CMB, on a straightforward mapping (nothing ad hoc) via only standard earth encodings and languages, then I think it's fair game to find it remarkable that the universe contains life.
(Nice to see you on Substack! I wonder if you remember an extensive correspondence we had over email in 2012 when you were in university. The topic was the MUH)
I completely agree that it's fair game to find it remarkable if the Quran (and no other text) is in the CMB, but - without meaning to be difficult! - I genuinely don't see the underlying principle which says that we should also find it remarkable that the universe contains life (allowing for the fact that there are some things we should find unremarkable, so there is some sorting principle at work).
I'm on board with your suggestion that this sorting must be in some way perspective-relative, but there's something more to unpack in the subjectivity of it, otherwise how would we reach agreement on what's remarkable and what's unremarkable?
On the face of it, what seems remarkable about the Quran case is that the text as it appears on Earth has a tremendously complex causal history with all sorts of contingencies, so to find the same text anywhere else at all with a non-overlapping history would defy probability.
But in the case of the universe being hospitable to life, it's not obvious to me that we're comparing multiple independent events which have a low chance of co-occurrence - are we? If it can be put that way, I would get it, otherwise maybe there's a better analogy to the Quran case.
At the moment, at least, the subjective pull towards accounting for the existence of life feels to me more like walking onto a beach full of pebbles, finding one I particularly like, and asking, "how did that get there?" I'm very open to changing my mind, though.
(Amazing that you remember a correspondence from 2012! Sad to say that I have terrible records and, hearing it now, I'm surprised I had opinions on the MUH! If I made sense, I'll be pleased; if not, I promise I've much improved!)
OK! Glad we agree about the Qur'an. So we have some common ground to build on.
I'm not claiming that you should find it remarkable that the universe contains life. But I do claim that most people do, and to be honest I suspect you do too.
If we had a random universe generator which would just generate universes with random constants and show us what they looked like, and 99.9999999999999% of the time it was just boring featureless voids or singularities of one sort or another, I find it difficult to imagine that you would not linger a little longer than usual to take a look if the computer printed out "Life detected". I find it plausible to think that you don't believe there is an objective principle by which you should find life interesting. I find it implausible to think that you actually don't find it interesting yourself.
I think that's enough. If you find it remarkable yourself, much more remarkable than you would expect to find if your observation were randomly selected, then I think you ought to be looking for an explanation. There is no need for any objective criterion of remarkability.
If you win the lottery yourself, then you should find that remarkable. You should investigate whether you might be dreaming, or in a Nozickian experience machine, or having a psychotic break, or the victim of a prank.
(And if the odds of winning are great enough, you probably should believe some such skeptical theory.)
But if you learn from the news that some randomer won the lottery, then that's not so surprising, as long as enough people play the lottery. So it's fine for what is surprising to be completely subjective. And as long as we find it remarkable that there is life in the universe, then we have something to explain.
If we need to put it on a pseudo-objective footing, there is a kind of coincidence to explain, if we go a bit meta. There is something observed by an observer who finds that observation to be extraordinarily significant. It's not a coincidence if I ask you what your birthday is, and it's just some insignificant date. It is if your birthday turns out to be the same as my own -- but that is only because that date is particularly significant to me. An observer has observed something that is of extraordinary significance to them. So there is something to explain. The birthday coincidence isn't really all that amazing, but you can imagine you might wonder if someone is playing a prank on you if we make the probability of the coincidence more extreme.
But I want to emphasise that the coincidence isn't really that we have the same birthdays. It's really more about what we find to be significant. If your birthday is that of my son's, then that's also a coincidence, but maybe less so if we assume that my son's birthday is of less significance to me than my own. If your current street address is somewhere I have lived, it's maybe more impressive if it's a more significant one, like the first address I ever had, or where I have lived the longest time, and less so if it's an AirBnB I stayed in for a week.
The parable of the pebbles is a good one. If you find a pebble you particularly like, the question is how unexpectedly significant is this pebble? If we think it is of astonishingly low probability that you would find a pebble that would strike you to this degree, then we do have something to explain, and we should ask "how did that get there?". But if it is not so suprising that you would be struck by a pebble to this degree, then there is nothing to explain.
(I remember this correspondence because it was a very long and fascinating one. I was arguing for the MUH, and you were giving me the most sustained and challenging opposition I think I had ever had for it. I have 61 emails from you in my inbox, from April to June of 2012. Your last email says...
> Sorry about the delay in response. Don't worry, I'm still interested in our conversation and will come back to your last email - I have just become considerably behind on my exam and it's due in 7 days, so I'm devoting all my efforts to it. I will get back to you soon. :)
The meta-discussion is good, though you might have to keep me on the right level: on one level, I can say that I do agree to the remarkableness of the apparent fine-tuning and I take it seriously by having this conversation! But within the conversation, I step down a level to do some analysis and I find that the only explanation I need to give is that I have subjective biases that sometimes misfire. Is this perhaps sufficient for you, or do you think we're all compelled to pick from the standard menu of responses?
Back on the Quran analogy, I left out a critical detail: when I asked whether we can talk about life in terms of a low probability coincidence, of course we could say that the coincidence is in the settings of the fundamental parameters of the universe. But *any* specific settings of the parameters would have a low co-occurrence probability, so we still need a prior interest in our settings.
In the analogy with the Quran, having a prior interest would be like saying that we care particularly about the Quran but we don't - the situation would hold of any human text discovered in the CMB. By that reasoning, I would absolutely concede that there is something remarkable in there being any settings of any parameters that lead to any universe, but that's a different argument.
So it seems to come down to our response to the random universe generator, and if I'm asked: "do you think you're more interesting than a void?" my answer is that that's worth thinking about but, ultimately, no - it's all just 'stuff' in different arrangements. Yes, I accept that, as the living kind of stuff, I notice a selfish difference, but I don't know how to distinguish this from spooky but meaningless coincidences.
Let me try it out this way. Suppose you were someone who never had dreams. For whatever reason, whenever you sleep it’s always nothingness. Then one night, you have a dream, and the next day your dream comes true. Are you going to attribute special significance to the dream?
I think this is closer to what’s being asked in the FTA, where the non-dream nights are the void universes, but while I think I would be more impressed by this circumstance, I also think I would dismiss it because I have so many other commitments in my worldview that would discourage a special interpretation.
So perhaps the more fundamental thing I’m saying is that I don’t accept that the FTA presents a self-contained argument. I think we’re all making different interpretations of it, as refracted through our other commitments, and mine tell me that nothing significant is being said. Maybe, as you said earlier, we just have to agree to part ways on this one, even though we otherwise agree on nearby cases like the Quran in the CMB.
(I'm all the more surprised about the 61 emails! Sorry that I didn't hold to my promise! Do you remember where the conversation started?)
> I find that the only explanation I need to give is that I have subjective biases that sometimes misfire.
I don't think that is sufficient for me, depending on what you mean.
What I'm about to say is a bit vexed by the impossibility of observing a universe which is not fine-tuned. So, again, lets imagine you're just randomly generating universes in simulation and having a look inside. That way we can focus just on the problem of whether it is interesting.
Suppose you are allowed to look at one universe, and it so happens to have life in it, and you find that interesting. You can chalk this up to a subjective bias and look no further for an explanation if and only if the bias is that you were likely to find any arbitrary universe so interesting and you are unaware of this fact. So, your bias might be that you are overestimating how interesting you find this particular universe as compared to possible alternatives. Maybe you would have been equally fascinated by a featureless void.
But, as long as you really are especially fascinated by this universe, and the probability of being so fascinated is extremely low, then there's an impressive coincidence, and we should be seeking explanations.
> In the analogy with the Quran, having a prior interest would be like saying that we care particularly about the Quran but we don't
I think you're misinterpreting the explanandum here. The explanandum is not why the constants have the particular values they have. As you say, any specific settings of the parameters would have a low co-occurrence probability. This is not what makes it interesting. What is interesting is that the parameters we have conspire to give the unvierse the property of being life-permitting. There are many other configurations that might also achieve this -- varying one constant in one way might be compensated for by varying another constant in another way. The problem is just that the vast majority of possible configurations do not appear to be life-permitting.
Focusing on the specific values of the constants reminds me again of the the pebble parable, just finding an arbitrary pebble and being amazed. No, what makes this interesting is more like finding a pebble of solid gold. It doesn't have to be any specific solid gold pebble, just any pebble with this very surprising property. It doesn't have to be the Quran. Any meaningful text would demand an explanation, but all the more so if it is a text with as much significance as the Quran.
> I don't know how to distinguish this from spooky but meaningless coincidences.
You don't have to. Any sufficiently impressive coincidence should send you looking for an explanation. Sometimes there are none. But sometimes there are.
For example, I just had a heating engineer over to replace a zone valve that had failed. It had failed months ago, and we just got around to fixing it, because we could make do with the other zone valve. All good. Then, that same night, I noticed that the other zone valve wasn't working. That's a coincidence! I still don't know if it's because the connection came loose because of the repair, or if it just failed randomly the same day. Either is possible, but the coincidence makes me suspicious that it had something to do with the repair. That's how this stuff should work. Coincidences should make you seek an explanation, but you're not guaranteed to find one. The more extreme the coincidence, the less you should be prepared to accept that it just is a coincidence.
For coincidences that are only brought to your attention because of how impressive they are, the explanation is straightforward. It's a big world. All kinds of random stuff happens. Sometimes coincidences happen, and you might hear about really impressive ones. So hearing about a really impressive coincidence *because* it is a really impressive coincidence should give you reason to think that it's a big world with all kinds of random stuff happening. Just like with fine-tuning (on my view). If some really impressive coincidence happens to you, personally, then you should consider whether there is a prank or you're dreaming or something.
> Are you going to attribute special significance to the dream?
Yes.
> while I think I would be more impressed by this circumstance, I also think I would dismiss it because I have so many other commitments in my worldview that would discourage a special interpretation.
That's fine. You're not guaranteed to find an explanation. But you should be trying to think of one. Maybe the dream wasn't as like what transpired as you think. Maybe some vague similarlity has grown in your memory, a bit like how people can read all kinds of stuff into the prophecies of Nostradamus. Or maybe it was a huge coincidence. But you should be impressed by the coincidence and you should be looking for explanations, in proportion to how impressive/improbable the coincidence is.
I think it’s helpful to boil down the talk of coincidences to two clearly stated independent things with a low co-occurrence probability. So, in our analogies, we have:
A. (1) an Earth text and (2) that text in the CMB
B. (1) the golden pebble and (2) that pebble on a regular beach
C. (1) I have my first dream and (2) that dream comes true
D. (1) a zone valve breaks and (2) that zone valve’s pair breaks
In each case, we speak casually about having a subjective impression of a coincidence in need of explanation, but that impression is a consequence of appreciating that (1) and (2) are objective facts with apparently non-overlapping causal histories.
Then the FTA comes along and, at least per your description above, it asks us to consider:
E. (1) a life-permitting universe exists and (2) I find that interesting
In this formulation, I’m essentially getting a NaN error because we’ve got an opinion instead of a fact for (2). If we leave it like that, I can say, sure, this is a neat topic for a coffee morning, but let’s not rush into updating our credences about metaphysical propositions like god and the multiverse.
Alternatively, we can try to amend the formulation, and probably the most sensible option is:
F. (1) the universe has variable parameters and (2) its parameters allow for life
For the sake of argument, we’re accepting the low probability of (2), but it doesn’t seem intelligible to speak of (1) having a probability at all, it’s just the context for anything that happens. So, then we’re just talking about (2) on its own and where has the coincidence gone?
Then we can say that it’s fine just to wonder about a single low probability event but to assess its significance we introduce subjective assessment, which is just (E) again and I’m back to NaN.
I agree that the multiverse response to fine-tuning is trash but it absolutely doesn't require something "Godish" to explain. Also I thought you said that explanation was not the right way to think about the issue?
I think it's fine to talk of 'explanation' if you want to make the point more accessibly. But when we get to evaluating disagreement, we should cut that out and move to Bayesian confirmation.
Do you think there's anything to the idea that, once we start reasoning in Bayesian terms, it may arguably be appropriate to treat knowledge of our own existence as background knowledge? I've already massively disconfirmed all versions of naturalism on which I don't exist in the way I appear to (as a biological organism). It seems to me that if I were to wake up in a white room, and I haven't gone and checked the physical parameters yet but I know I'm a biological organism, then I'm going to reason like this: if naturalism is true, I will almost certainly observe life-permitting parameters, but if theism is true, I might not observe life-permitting parameters since God could've supernaturally created me in this room. Then I go and observe life-permitting parameters. In that framing where your knowledge of your own existence is already part of the background, the data seems to support naturalism. Some creationists like to argue that abiogenesis is so implausible that a supernatural creator would need to intervene in the process to produce living cells. So it seems like at least some theists would agree that if we found ourselves in a non-life-permitting universe, that would be evidence for theism and miraculous intervention. But it can't be that finding yourself in a life-permitting universe is evidence for theism and finding yourself in a non-life-permitting universe is also evidence for theism, it would just be that life-permittingness is evidentially neutral towards theism. And then what you're left with is no longer fine-tuning data, it's just the fact that you exist.
Sorry my first reply was horribly worded. I'll try again: Jane's situation doesn't require an explanation. You cannot 'explain' pure statistical probabilities in the same way you can look for explanations for other unusual events that happen in life. When you win the lottery, you don't hand back your winning ticket and refuse to cash it because it clearly couldn't have happened as the odds were so against it. Statistical probabilities are literally not things we can look for scientific explanations for. The are a different category of things, to things such as everyday physical observations.
In the same way the multiverse people are very misguided to think they need to explain life in the universe, the same is true for the fine tuning argument. Multiversers are forced to posit an "infinity" of multiverses in their multiverse theory, because anything less than infinity mutliverses creates glaring logical and statisitical issues in the theory (namely - each individual universe starts with the exact same odds of life, not more or less). But they need'nt bother, there's no required explanation for merely a large statistical probability.
Similarly, fine-tuners think that the parameters of the universe combined with the existence of life "requires" some kind of rational explanation, and I'm not sure why. You can't get a rational scientific explanation of why you won the lottery, or why your sperm beat millions to reach the egg. Your personal phsical existence in time now is already SO unlikely in regards to pure mathematical odds we don't need to appeal to fine tuning of the universe, we've *already reached* numbers that we usually accept as being *physically impossible* just by being born. So that's why we cannot use statistical probabilties in this way. It's a misunderstanding of statistics. Incidentally there's some good modern philosophy out there on this subject of statistics!
Explanation has to stop somewhere. This might be a good worry if you responding to some kind of 'Why is there something rather than nothing' argument, but it's irrelevant to a Bayesian argument.
I see many people who get good replies to their posts but ignore them even if they ought to have the time to reply to them. I'm not talking about someone who have received hundreds of replies. But people who get 5-10 good replies on one of their two posts that day, yet ignore to answer any of them. You seem to be the exact opposite. Really nice to see for a change someone who doesn't come of as elitist.
The crowd-summoning analogy is interesting, but it presupposes a pre-existing pool of possible observers who are assigned to an observed success. Structurally, that is much closer to a Cartesian-dualist or SIA-like setup than to a physicalist one. In my preprint, I argue that in just such a case one indeed gets confirmation of a multiverse over a single-universe hypothesis independently of fine-tuning.
The Jane case, by contrast, is much closer to physicalism. Here, observers are themselves outcomes of universes, not pre-existing candidates for assignment. In that case, if one includes the total evidence while still accounting for selection effects, the calculation shows that there is still Bayesian confirmation of a multiverse over a single-universe hypothesis, but it remains dependent on fine-tuning. See my preprint for the calculation: https://philpapers.org/rec/NABTMV
On the first point, I think you're taking the analogy a bit too literally. It's just a way of capturing the improbability of me in particular existing, as opposed to some other possible creature.
I still haven't had chance to read this, I'm afraid. Maybe check out the maths Kenny Boyce has done for the paper we're working on together (he out the basic maths on his webstie) and see where we part company.
What I’m taking literally is the probabilistic structure: (i) a fixed pool of possible observers, and (ii) any of them can end up in one of the unlikely scenarios. From a probabilistic point of view, that is exactly an SIA-style setup: a fixed pool of possible observers who can end up in any universe. Structurally, that corresponds to a Cartesian-dualist scenario. So it should not be surprising that the crowd-summoning case yields the same result as SIA: confirmation of a multiverse over a single-universe hypothesis independently of fine-tuning. I therefore do not dispute Kenny’s calculation; my point is that this setup already builds in the stronger self-locating assumptions.
The Jane 3 case is different, because there the observer is an outcome of the process, so the pool of observers depends on the available processes. If the evidential claim is that someone with my evidence exists — the issue discussed inconclusively in the video around 1:14, where Kenny says he’d “like to see that build out a bit more” — then, on my calculation, this still yields confirmation of a multiverse over a single universe, but now in a way that remains dependent on fine-tuning.
I would argue that the fine-tuning of our universe gives us evidence for the multiverse (with varying constants) regardless of the fact that the multiverse hypothesis can't explain why our universe, in particular, is fine-tuned.
First, we have to keep in my mind that the surprising fact (the thing that's begging for an explanation) about living in a fine-tuned universe is that the universe is fine-tuned for LIFE and NOT that it's fine-tuned for US specifically. The universe being fine-tuned for life doesn't necessarily entail that it's fine-tuned for us (in the sense of allowing us to potentially exist in it, which also includes the fact that we can only exist in one specific universe).
This already makes the fine-tuning case non-analogous to Jane. What Jane is trying to explain is HER fertilization, not the fact of fertilization itself. It's the latter case that would be analogous to the case of fine-tuning for life, where the thing to be explained is the existence of fine-tuning for LIFE, not US specifically. So, the real question is: by being born and knowing that the odds of fertilization of any given egg are low, does Jane get support for the hypothesis that there must be multiple eggs out there assuming that she's not trying to explain the odds of HER egg getting fertilized but the fact of fertilization itself?
But what about the requirement of total evidence? Shouldn't we include the fact that OUR universe is fine-tuned when evaluating competing hypotheses (multiverse or theism)? Yes, but note that when we include that piece of evidence we're no longer making inferences just from living in a universe fine-tuned for LIFE but also from living in a universe fine-tuned for US. If we're only considering the fact that we live in a universe fine-tuned for life, we can get to the multiverse (the fact that it happens to be our universe that's fine-tuned is neither here nor there), but if we add the fact of living in universe fine-tuned for us, the multiverse seems like a poor explanation.
However, I don't think single-universe theism fares any better than the multiverse! We have to change a couple assumptions in our reasoning to see this though.
If I'm not mistaken, the inverse gambler's fallacy charge against the multiverse hypothesis goes something like this: My total evidence is that I live in a universe fine-tuned for life and, GIVEN THAT MY UNIVERSE EXISTS, a multiverse with randomly varying constants does not give a high probability for MY universe being fine-tuned for life. No matter how many universes there are, the chances of mine being fine-tuned for life are low whereas single-universe theism gives way better odds.
However, the multiverse theorist makes different assumptions. What seems surprising to them about living in a universe fine-tuned for life is the existence of fine-tuning for LIFE, NOT that THEIR universe specifically is fine-tuned. They will then reason like this: If there is a multiverse, the probability of the existence of fine-tuning for life is high and, GIVEN THAT I EXIST, the fact that my universe is fine-tuned is not surprising either (this is the selection effect).
It looks to me like taking MY existence vs MY UNIVERSE'S existence as given in our reasoning leads to radically different conclusions here (in addition to the fine-tuning for life vs us difference). If I take the existence of my universe as given, it seems surprising that it's fine-tuned for life, but if I take my existence as given, that surprise dissipates. The latter supports the multiverse and single-universe theism whereas the former only supports single-universe theism.
If we drop both givens and try to explain why our universe both exists and is fine-tuned for life (therefore also allowing us to exist), both hypotheses, the multiverse and single-universe theism, fare quite badly. The multiverse arguably gives our universe higher chances of existing but fails to explain why it's fine-tuned. Single-universe theism explains why our universe is fine-tuned but not why it exists. It doesn't seem likely that out of the infinite amount of possible fine-tuned universes (with universe-specific observers) that God could have created, he decided to create ours (and us) specifically.
So, perhaps multiverse theism fares the best? I don't know.
To conclude:
1) We CAN get from the existence of fine-tuning for life to the multiverse. It's a logical and legitimate inference to make. However, if we include both pieces of evidence, that fine-tuning for life exists and that our universe is fine-tuned for life, the multiverse is not a great explanation.
2) The inverse gambler's fallacy charge works only if we merge the two concepts (fine-tuning for life vs our universe's fine-tuning) and we take the existence of our universe as given and try to reason from its being fine-tuned to the multiverse. If we keep them separate and take our existence as given, the multiverse hypothesis is solid because there's only one surprising fact to be explained, the existence of fine-tuning for life.
3) If we try to explain why our universe exists AND is fine-tuned for life, the multiverse and single-universe theism are equal(ly bad) explanations.
(And yes, I created a substack account just to reply to this...)
Jane seems disanalogous to me because the doctor scenario is just bonkers and presumes she is an immaterial spirit that can drop into any egg. Multiverse, on the other hand is a hypothesis that predates fine tuning and for which there are independent reasons to believe.
As someone who grew up around Eastern ideas of karma and purpose, I find your work on cosmic fine-tuning strangely familiar.
The question “Is this all an accident, or is there a telos?” appears both in physics and in people’s private narratives about their lives.
What I appreciate here is your willingness to sit with that question analytically, without collapsing into either dogmatic design or nihilistic randomness.
Sean Carroll makes this kind of argument, but I find it pretty weak. Suppose God had created the universe with other constants, and had just allowed planets to form and the sun to give energy by magic.
However things end up working, they're going to have to work in some logically consistent way. There are going to be regularities. We can do science on it. The net effect would be that the apparent laws of physics would look different than the present laws of physics but with life-incompatible constants. It's going to look suspiciously like everything is fine-tuned to support life. Much like this universe, perhaps.
There are other responses too. Another one would be just to go back to the "why not?". If God can do anything, then all options are equally easy, so God may as well pick from them according to criteria that may as well be aesthetic. Why think that God prefers a magic aesthetic where there's a sort of Rube Goldberg machine set up to allow life to exist despite the basic incompatibility of the rest of the laws of physics as opposed to the elegance of coming up with one set of laws that just conspires to allow life to evolve?
Philip's own response is just that it's a Godish thing of limited powers. The structure of the laws of physics is set somehow. It's metaphysically or logically necessary, perhaps. He doesn't say or know where they come from. But it's just the same problem for him as it is for naturalism, so it's not worse. In Philip's hypothesis, God only has the freedom to set the constants, and if that is the case, there are very few possible tunings that allow life to exist, so those are the values God chooses.
I agree mostly, I think. But then the god that is most popular has created non-physical life, presumably not bound by those restrictions. Are Heaven and Hell logically consistent? Why create a physical realm at all?
Given the statistics of the tuning, asking why not? just prompts me to just say sure... but why tho? If it needs an explanation, it needs an explanation...
Then there's the limited power god, well OK but what makes it a god? All bets are off in that case, why assume he has control of the parameters? They could be fixed by other laws. This being could be an advanced scientist who created our universe entirely by mistake. Or maybe not a mistake, maybe he made a Doomsday weapon to destroy his universe by replacing it with ours. The possibilities could be endless.
May have missed your points entirely, just thinking out loud at this point 🙂
I'd just say that fine-tuning is (arguably) evidence for something Godish, not necessarily for the most popular God. It may not be very strong evidence for any particular account, but it may be evidence against the default assumption that there is just one, naturalistic universe. So, yeah, God could be a scientist who created a universe in a lab, or God could be some teenager in a basement with a supercomputer capable of simulating a universe. Philip has independent reasons for favouring his account over others. Fine-tuning alone won't cut it.
True, and I'm pretty sure I wouldn't be convinced by the other reasons either (moral objectivity I'm assuming).
If I were a god of limited power, who just lights the blue touch paper and stands back, I would want to give my baby the best chance of developing as I want, so in choosing the parameters I would more likely pick a set of options that did so more often than not. Seems safer than what we appear to have. What if I got one of the numbers ever so very slightly wrong?
Well, Philip’s idea is that the structure of the laws are fixed and God(ish) only has the ablility to change the parameters. It’s the nature of the structure of the laws that makes the existence of life so precarious. God(ish) couldn’t have done much to make life more probable than it is, within those constraints.
So Philip is assuming that Mr. God(ish) can only alter the parameters of the physical laws, not the structure... sounds a bit like Philip is fine-tuning God..?
This is the difference between a homogeneous multiverse and a heterogeneous multiverse, right? God(ish) can only choose from the options on the menu of the homogeneous one, for some reason.
Jane is right: her existence is evidence for the existence of the relevant class of observer-producing outcomes, and that class is more likely under a multiverse hypothesis than under a single-universe hypothesis. So a multiverse can be confirmed even if the probability of *her* existence is no higher.
I think what's doing a lot of the work in the Jane case is that intuitions are being pumped by the weirdness of your situation.
Unless you posit that everyone is being made by crazy doctors, your situation is a very weird one, even if there are lots of crazy doctors. Being made by a crazy doctor is not a precondition for an existence of someone like you. Unlike fine-tuning.
So there's kind of a reference class problem. If you take the most intrinsically significant thing about your identity to be that you were made by a crazy doctor, then you might want to posit lots of crazy doctors so that somebody like you can exist. But if you take this fact about your origin to be insignficant and incidental to your identity, then you'll be inclined to disregard it and see it as something that doesn't need an explanation.
In the actual fine-tuning case, what is salient or significant is that conscious observers and life exists at all. That does not seem to be so amenable to treatment as an insigificant feature of the universe that only matters to us because it happened to take place, though some opponents of the FTA want to say so. We look at the evidence for fine-tuning, and it strikes us that this is really important. That a universe without fine-tuning would be utterly different. That this is really important and demands an explanation.
That you were made by a crazy doctor? Not so much. It's weird, but it doesn't strike us as all that significant. It doesn't demand an explanation in the same way. It's more like being born in an aeroplane, or being conceived as a consequence of your parent going on a holiday they could only afford because they won the lottery. There are all sorts of stories about how people are conceived and born, and they're all unique. Nothing about Jane really cries out for an explanation in particular. Every conception is a miracle. Jane's is no different to anybody else's.
But if Jane doesn't know if there are other humans, and if Jane thinks that humans can only be created by crazy doctors, then she may have reason to think there are many crazy doctors. Or if Jane thinks of people who are made by crazy doctors as being importantly different from regular people, as different as we are from bacteria, then she might think that there are lots of crazy doctors.
My other thoughts on this video are here: https://disagreeableme.substack.com/p/the-fine-tuning-argument-is-tough
Interesting! How would this figure in the Bayesian calculation? Different prior? If so, then you still ought to get an evidential boost for Many Doctors, but that doesn't seem right.
This is exactly right. Philip would object that the existence of other crazy doctors does not increase the probability of HER existence. But when we compare physical theories, the relevant evidence is class evidence, not token evidence. So the right perspective is the “eye-in-the-sky” one:
"Alan is a spy who wants to find out how many doctors work at the IVF institute. Alan, like everyone else, knows that every doctor in the country has a strange disorder: each doctor performs at most one IVF procedure, and only if they roll double six on the first try. Alan waits in the bushes near the institute for several years. He never sees the doctors, because they use a secret tunnel. The only information available to him is whether babies are brought out. One day, Alan observes that a baby is brought out—Jane. He reasons as follows: if there is only one doctor, the probability that a baby is born is 1/36; if there are many doctors, the probability is much higher (approaching 1 as the number of doctors increases). Hence, given that a baby is brought out, it is more likely that there are many doctors." (https://philpapers.org/rec/NABTMV)
It (life in the universe) does not “demand an explanation” at all. This is the fundamental statistical error of your entire argument. Giant statistical probabilities do not come with ‘explanations’. They are just what they are. You cannot use statistics in this way and this is what people just don't understand.
I disagree.
If you win the lottery, then, make sure you return the ticket because it was clearly statistically impossible to any reasonable measure that you just won, and that you demand a logical explanation for it before accepting your winnings.
Thank you for illustrating your point so I can respond.
When I say something "demands an explanation", I just mean you should seek an explanation. You're not guaranteed to find one.
"Statistically impossible" just means improbable. Winning the lottery is very very improbable. If you appear to win the lottery, you *should* be asking yourself if winning the lottery is really the best explanation for your evidence. You should be wondering if you're dreaming or the victim of a prank. You should, in other words, be looking for an explanation.
If one is not forthcoming, then eventually you just have to accept that the best explanation of the evidence is that you have actually won the lottery. But that's only because there is no better explanation. It is never the right attitude to just ascribe some very unlikely event to a freak coincidence when alternative explanations are more likely.
Your problem is that you're basing your beliefs upon this statistical idea you've discovered (the universes precise parameters). But like I'm trying to explain, these really big statistical measures literally do not work in the way you think they do. You can't even say you 'should' look for an explanation. There is very good modern analytical philosophy on statistics I could recommend you because many people, even philosophers make this mistake.
In every day life, if, say, the 80 year old president of a country suddenly does a speech in Polish, having shown no prior evidence of knowing our speaking Polish, then that is a unusual event that we should, and could, look for an explanation for. There would also be an explanation we could find. But this rare and odd event is *not* analogous to modern statistical measurements (of gigantic numbers), for evident reasons that can be explained. Statistics are not giving you the same type of information. The fine tuning argument is the same thing. It's not giving you the evidence for something rare that you think it is. It's purely statistical. You can't use this as actual evidence of anything.
I do not agree that there is a relevant difference between spotting something about a numerical distribution and spotting something odd about a president speaking a language.
What exactly is the difference you’re trying to point to? What criterion are you using? Is it just that numbers are involved? If I roll a die 10 times and get only sixes, is that evidence that the die is not fair?
I think that this is correct: what's doing at least some of the work here is that our background prior on "multitude of crazy doctors rolling dice" is so low compared to our prior on the "the doctor cheated/dice were rigged" hypothesis because rigged dice are relatively common.
But if Jane does some investigating and finds that these were, idk, quantum dice, such that rigging the dice would have required someone to modify the most fundamental forces of reality, then surely this does give Jane evidence for the many doctors hypothesis over the "cheating" hypothesis
My intuition is that it doesn't provide any boost at all. But if you don't share that intuition, we have to turn to the broader analysis of what's going on in different cases, that Kenny Boyce, Mark and I has been doing
When you say 'teleological universe', how would this fit, or not, with Lee Smolin's and Rupert Sheldrake's (as best I know independent) proposals that natural laws evolve? Does the fine tuning depend on this being a tunable instrument? And if so, do we find an example of self-tuning in homeostatic organisms?
I didn't mean Smolin's story, but that would be a third alternative i haven't considered here.
I love the Jane story.
I have an additional issue with the multiverse theory. As an argument against ‘something godish’ it has the same weakness as the argument that the laws of nature, or for that matter evolution, dispenses with the need for something godish.
Seventeenth century scientists developed the idea that the laws of nature are the regularities with which God governs the world. By the middle of the eighteenth century some people including David Hume could take for granted that the laws of nature are self-explanatory, and therefore debate whether there is a god at all. But Hume spotted the problem. As an empiricist he noted that we never observe causation. We infer it. The laws of nature are the regularities, not the causes of the regularities. Once this point is seen it can be applied to evolution and multiverses just as well. Any or all of the three could be God’s way of creating stuff. If they aren’t, we still have no account of causation, just a map of successive events.
Thanks Jonathan, i"m definitely not a fan of Humeanism!
This assumes a Humean view where laws are merely descriptive summaries of what happens. That is highly controversial. Frameworks like Constructor Theory and Modal Realism suggest instead that laws are grounded in underlying governing principles. If these principles are the inherent architecture of a concrete world, they are the cause of the regularities.
How does this help? If the laws of nature are the inherent architecture of a concrete world, how does calling them the inherent architecture make any difference? (And why should Hume disagree?) We can still ask what caused them to be there. In other words, how have you done more than just add another layer of definition?
It helps in the sense that it reduces physical laws to ontological, logical necessities. Modal Realism and Constructor Theory propose that the 'architecture'—the underlying laws of what is possible and impossible—is the fundamental floor. At some point, one hits the 'Brute Fact' of logical and physical consistency, which is a far more robust explanation for reality than Hume’s 'infinite coincidences' or the necessity of a creator.
So the ‘architecture’ is the ‘fundamental floor’. It relates to the physical laws by being different from them and being the cause of them. As you say this is a theory, not accessible to empirical evidence. Use of the term ‘floor’ suggests to me that the theory denies that it is in turn caused by something else. It sounds to me like an ‘uncaused cause’. If I have understood this correctly, I’d like to ask how it differs from Thomas Aquinas’ uncaused cause, namely God. You say it is ‘far more robust’ than either Hume or ‘the necessity of a creator’. How is it more robust? For Aquinas the uncaused cause was an agent, more intelligent than humans, deciding to create. To enquire further we must reflect on the nature of God, which people have done all through history. Your fundamental floor, as I understand it, is ‘just there’. To me this conjures up a picture of an implied noticeboard forbidding us to enquire any further. Isn’t that rather unscientific? Or is there more to it?
While Modal Realism is metaphysical, Constructor Theory provides a framework for potentially testable predictions. In both cases, the 'fundamental floor' is logical necessity. A purported deity either equates with logic or is subject to it. In the first case, this is merely ontological relabeling. In the second, it introduces an extra entity not required by the theory, and is therefore, by Occam’s razor, less robust.
Furthermore, I’m curious about your stance on the nature of the 'Uncaused Cause.' In Aquinas’s time, little was known about the mechanics of intelligence, but modern science (neuroscience, AI) reveals that intelligence is a product of immense functional complexity. If you hypothesize a 'highly intelligent' deity, do you regard them as a complex entity or a monadic, simple one?
If the former, we must ask what 'architecture' allows for that complexity; if the latter, you are positing a type of intelligence that contradicts everything we currently understand about information processing and requires a nomologically radical hypothesis—making it far less robust than nomologically conservative theories like Constructor Theory or Modal Realism.
Thank you. I find this fascinating and am happy to continue the dialogue, if you are willing. But please ignore me if I’ve passed my Best Before date.
So the ‘fundamental floor’ is logical necessity. Does this mean that what is logically impossible cannot exist? I think this exaggerates the power of human logic. Lots of theories have proved true even though they defied logic at the time: wave v particle theories of light, the Einstein-Podolski-Rosen experiments with electrons, etc. I’m sure you can think of others. At the other extreme there are theorists, like the evolutionary psychologist Patricia Churchland, arguing that the human mind has only evolved for the four ‘f’s (feeding, fighting, fleeing and reproducing) so ‘Truth, whatever that is, takes the hindmost’. (I’m not defending her. It’s just that your position sounds to me like the other extreme.) So when you say ‘A purported deity either equates with logic or is subject to it’ you seem to me to be denying that something can exist which transcends human understanding. For the purposes of empirical research based on the scientific method this is usually good enough – but not for all purposes.
On Ockham’s razor. It’s a useful heuristic tool, but only a tool. Everything from the molecules in Ockham’s fingernails to the Milky Way has turned out to be far more complex than anyone knew in his day. The simplest solution isn’t always the right one, to say the least. So when you speak of ‘extra’ entities ‘not required by the theory’, research doesn’t limit itself to the current theories, at least not in the long term.
On divine creators. They have been hypothesised since prehistoric times in every human society as far as anthropologists have observed. So when you describe them as a ‘nomologically radical hypothesis’ I wonder what you mean. Your ‘conservative theory’, as a replacement, only goes back to eighteenth century France. More important, people have believed in divine creators for other reasons too – but that would take us onto other topics.
I wonder whether the underlying difference between us is that where you respond to the prospect of a divine intelligence by asking what ‘fundamental floor’ could allow for it, I’m responding to the prospect of a ‘fundamental floor’ by asking what intelligence could allow for it. In a sense your foundation is stuff, mine is intelligence. Does that sound right to you?
I don't see that as a problem with the multiverse explanation. That's a separate problem. The problem of where the laws come from, or why anything exists. The multiverse is only posited to explain fine-tuning. It doesn't have to explain everything. If you want to take the existence of the laws as evidence for God, that's fair game. But it would probably invite different responses from the atheist side (which would likely be compatible with the multiverse explanation).
I don't think an appeal to the anthropic principle actually requires multiple universes?
The basic argument is that if the universe wasn't fine tuned for life, we wouldn't be here to discuss it, so we can't infer anything about design from the fact that the universe is fine-tuned. Whether life was extremely improbable (fluke occurance) or guaranteed in some universe (multiverse), the argument still works.
It would be like if Dr. Smith later told Jane this story, and didn't mention what the dice rolls were, just mentioned that if it hadn't been all sixes, Jane wouldn't have been born. Jane can infer from this that Dr. Smith must have rolled all sixes, even without asking.
Your earlier Scripture in the Big Bang argument is a lot more interesting in my opinion, since that would be a fact about the universe which has no obvious connection to the existence of intelligent life.
The real question is: Can we come up with any way the universe is "fine-tuned" that isn't contingent on the same things that intelligent life is contingent on? Possibly.
There are different versions of the anthropic principle, but as far as I can tell most people do not find the kind of explanation you are offering to be satisfactory.
If the universe wasn’t fine-tuned, we wouldn’t be here to discuss it. This means we know the universe must be fine-tuned. It doesn’t mean that this is unsurprising.
If I tell you that I am secretly an alter ego of Donald Trump, then (assuming you believe me), you know that I must be the president of the USA. It doesn’t mean that this is unsurprising. Why would Donald Trump be posting on substack about philosophy? It doesn’t seem like something he would do.
Let me try to rephrase it in terms of statistics:
The research question: Are universes fine tuned for life?
The sample size is 1.
The sample is biased because it wouldn't exist unless its universe was fine tuned for life.
irrelevant hypothetical a-priori chance a universe is fine tuned for life: 1 in 100^100
The result: The universe is fine tuned.
Since the sample size is 1 and the result is 1 out of 1, the p-value would be 1 in 100^100. Except the problem: The sample is biased. Due to this bias, to be intellectually honest, we have to throw out the only data point. This leaves us without a p-value.
The multiverse or not multiverse question is a question of population size, not sample size, so it really has little statistical relevance here. Unless you are trying to be *extremely* precise, you don't need to know the population size to do statistics.
A biased sample does not force us to discard the data point. It forces us to model the selection effect properly, so that the evidence can still function as objective evidence.
The multiverse question is not about population size. It is about which hypothesis makes the existence of the relevant kind of observer-containing, fine-tuned universe more probable. Once the evidential setup is reconstructed in a modal, third-person way, the result is still confirmation of a multiverse over a single-universe hypothesis.
I don't think that is the research question. The research question is "Why is this universe fine-tuned for life?"
We don't know the population size. So we're debating whether we can explain it by a biased sample from a large population which is bound to have some universe fine-tuned for life, or as an unbiased sample.
You and I both think the sample is relevantly biased. But that is just the point of contention.
Philip and Kenny essentially think that this is the only universe we could have sampled, so the choice of universe is not biased. I think that because only fine-tuned universes can be sampled, then it is biased. Which way you look at it depends on whether you're more focused on who exactly is doing the sampling or on what kind of thing can be sampled.
Then Philip and Kenny assume for the sake of argument that we could have sampled other universes. But then their argument shows that fine-tuning has nothing to do with the inference to a multiverse. No matter how fragile or robust the conditions for life to be possible, we're more likely to be sampling something the more universes there are.
This is a reply to both you and Richard.
The question I am asking isn't whether a single universe or multiverse is more likely. You're right that the unlikelihood of fine tuning is a good argument for the multiverse as compared to a single fluke universe. I agree with you and don't contest that point. That is not my point.
What I am contesting is the idea that a single universe that happens to be fine tuned implies design. In my opinion, it doesn't imply design, assuming that the fine tuning can be shown to be a per-requisite for anyone asking this question in the first place.
If I understand you correctly, your view is that since we exist, the observational datum already guarantees that the universe is fine-tuned, so the probability of a fine-tuned single universe should be treated as 1. But that seems to conflate epistemic reasoning with ontological reasoning: from the fact that we now know we exist, it does not follow that a fine-tuned universe was not antecedently improbable. It only follows that our evidence is conditional on our existence.
That said, I am sympathetic to a modal-realist-like metaphysical view on which fine-tuning needs no further explanation, but that is beside the point here.
I don't think I agree with you on that point.
If the Qur'an were written out in the cosmic microwave background, that would imply design, even if that being written out in the cosmic microwave background would be a pre-requisite for anyone asking "Hey, how come the Qur'an is written out in the microwave background?".
how could that possibly be a pre-requisite? Certainly such a universe would not be the only possible universe.
I think it's actually even worse than that. If what we mean by "life-permitting" amounts to the physical parameters being such that stars can form and life can arise in a way that doesn't violate laws of physics, then on omni-theism, God *could* create life much like us via miraculous intervention in a non-life-permitting universe. He doesn't actually need a fine-tuned/life-permitting universe to create life like us. So, yes if naturalism is true and life exists, it's nearly a guarantee the universe is "life permitting." But if omni-theism is true, this is not a guarantee because we may be here via miraculous intervention (contra physical law).
Here's the question: Forget all this chemistry and biology giving rise to life.
Can some type of non-living consciousness exist in a chaotic and disordered universe?
If it can, the real mystery is why the universe is so ordered and intelligible (and if Einstein couldn't answer this question, then probably I can't either!)
The anthropic principle sortof explains why consciousness must exist, but it doesn't explain why reality isn't a never-ending acid trip.
For the fine-tuning argument, the relevant reference class is not just “conscious beings,” but conscious epistemic systems capable of discovering that they inhabit a fine-tuned universe and reasoning about it—otherwise we would not be having this discussion in the first place. In that sense, radically chaotic or unintelligible universes are not relevant counterexamples here, because they would not support the kind of epistemic system whose evidence is at issue.
Assuming some version of origin essentialism is true, God couldn't create us in a different universe. Our origins is necessarily tied to who we are. So if God wants to actualize us, He has to fine tune the universe. And this is also why we couldn't have existed in another universe.
I would note that in the comment you're replying to I'm saying "life much like us" not "us" simpliciter.
I've never understood the premise that life is in special need of explanation any more than if the universe contained nothing. The anthropic principle strikes me as ironic in that, at the same time as it tries to make humans non-special (we're on one of many planets, in one of many universes), it begins with the anthropocentric assumption that humans need a special accounting.
But while the Many Doctors Hypothesis is clearly wrong, isn't it because Jane just shouldn't be asking the question about the improbability of her birth, which, retrospectively, was inevitable? Wouldn't the analogy to the fine-tuning argument for god suggest that Jane should think the dice were rigged?
I think it's enough for the coincidence to be subjective. The idea that this sort of coincidence has to be objective to warrant explaining is a mistake, in my view.
Otherwise, what you're saying suggests you would find nothing remarkable or in need of explanation in finding the Qur'an in the cosmic microwave background.
You could say that the fact that it is meaningful to us is just an artifact of the fact that we happen to have an encoding, a script and language on Earth (Unicode, Arabic script and language) that makes it seem meaningful. To an alien race with no such script or language the markings would look like random noise.
So there is nothing objectively remarkable or in need of explanation about finding the Qur'an clearly spelled out in the CMB.
If that's what you think, then I guess we'll just have to part ways, because this is clearly nuts to me. It's a wilful disregarding of very clear evidence.
On the other hand, if it is fair game to find it remarkable that the Qur'an is clearly in the CMB, on a straightforward mapping (nothing ad hoc) via only standard earth encodings and languages, then I think it's fair game to find it remarkable that the universe contains life.
(Nice to see you on Substack! I wonder if you remember an extensive correspondence we had over email in 2012 when you were in university. The topic was the MUH)
I completely agree that it's fair game to find it remarkable if the Quran (and no other text) is in the CMB, but - without meaning to be difficult! - I genuinely don't see the underlying principle which says that we should also find it remarkable that the universe contains life (allowing for the fact that there are some things we should find unremarkable, so there is some sorting principle at work).
I'm on board with your suggestion that this sorting must be in some way perspective-relative, but there's something more to unpack in the subjectivity of it, otherwise how would we reach agreement on what's remarkable and what's unremarkable?
On the face of it, what seems remarkable about the Quran case is that the text as it appears on Earth has a tremendously complex causal history with all sorts of contingencies, so to find the same text anywhere else at all with a non-overlapping history would defy probability.
But in the case of the universe being hospitable to life, it's not obvious to me that we're comparing multiple independent events which have a low chance of co-occurrence - are we? If it can be put that way, I would get it, otherwise maybe there's a better analogy to the Quran case.
At the moment, at least, the subjective pull towards accounting for the existence of life feels to me more like walking onto a beach full of pebbles, finding one I particularly like, and asking, "how did that get there?" I'm very open to changing my mind, though.
(Amazing that you remember a correspondence from 2012! Sad to say that I have terrible records and, hearing it now, I'm surprised I had opinions on the MUH! If I made sense, I'll be pleased; if not, I promise I've much improved!)
OK! Glad we agree about the Qur'an. So we have some common ground to build on.
I'm not claiming that you should find it remarkable that the universe contains life. But I do claim that most people do, and to be honest I suspect you do too.
If we had a random universe generator which would just generate universes with random constants and show us what they looked like, and 99.9999999999999% of the time it was just boring featureless voids or singularities of one sort or another, I find it difficult to imagine that you would not linger a little longer than usual to take a look if the computer printed out "Life detected". I find it plausible to think that you don't believe there is an objective principle by which you should find life interesting. I find it implausible to think that you actually don't find it interesting yourself.
I think that's enough. If you find it remarkable yourself, much more remarkable than you would expect to find if your observation were randomly selected, then I think you ought to be looking for an explanation. There is no need for any objective criterion of remarkability.
If you win the lottery yourself, then you should find that remarkable. You should investigate whether you might be dreaming, or in a Nozickian experience machine, or having a psychotic break, or the victim of a prank.
(And if the odds of winning are great enough, you probably should believe some such skeptical theory.)
But if you learn from the news that some randomer won the lottery, then that's not so surprising, as long as enough people play the lottery. So it's fine for what is surprising to be completely subjective. And as long as we find it remarkable that there is life in the universe, then we have something to explain.
If we need to put it on a pseudo-objective footing, there is a kind of coincidence to explain, if we go a bit meta. There is something observed by an observer who finds that observation to be extraordinarily significant. It's not a coincidence if I ask you what your birthday is, and it's just some insignificant date. It is if your birthday turns out to be the same as my own -- but that is only because that date is particularly significant to me. An observer has observed something that is of extraordinary significance to them. So there is something to explain. The birthday coincidence isn't really all that amazing, but you can imagine you might wonder if someone is playing a prank on you if we make the probability of the coincidence more extreme.
But I want to emphasise that the coincidence isn't really that we have the same birthdays. It's really more about what we find to be significant. If your birthday is that of my son's, then that's also a coincidence, but maybe less so if we assume that my son's birthday is of less significance to me than my own. If your current street address is somewhere I have lived, it's maybe more impressive if it's a more significant one, like the first address I ever had, or where I have lived the longest time, and less so if it's an AirBnB I stayed in for a week.
The parable of the pebbles is a good one. If you find a pebble you particularly like, the question is how unexpectedly significant is this pebble? If we think it is of astonishingly low probability that you would find a pebble that would strike you to this degree, then we do have something to explain, and we should ask "how did that get there?". But if it is not so suprising that you would be struck by a pebble to this degree, then there is nothing to explain.
(I remember this correspondence because it was a very long and fascinating one. I was arguing for the MUH, and you were giving me the most sustained and challenging opposition I think I had ever had for it. I have 61 emails from you in my inbox, from April to June of 2012. Your last email says...
> Sorry about the delay in response. Don't worry, I'm still interested in our conversation and will come back to your last email - I have just become considerably behind on my exam and it's due in 7 days, so I'm devoting all my efforts to it. I will get back to you soon. :)
Better late than never!
)
The meta-discussion is good, though you might have to keep me on the right level: on one level, I can say that I do agree to the remarkableness of the apparent fine-tuning and I take it seriously by having this conversation! But within the conversation, I step down a level to do some analysis and I find that the only explanation I need to give is that I have subjective biases that sometimes misfire. Is this perhaps sufficient for you, or do you think we're all compelled to pick from the standard menu of responses?
Back on the Quran analogy, I left out a critical detail: when I asked whether we can talk about life in terms of a low probability coincidence, of course we could say that the coincidence is in the settings of the fundamental parameters of the universe. But *any* specific settings of the parameters would have a low co-occurrence probability, so we still need a prior interest in our settings.
In the analogy with the Quran, having a prior interest would be like saying that we care particularly about the Quran but we don't - the situation would hold of any human text discovered in the CMB. By that reasoning, I would absolutely concede that there is something remarkable in there being any settings of any parameters that lead to any universe, but that's a different argument.
So it seems to come down to our response to the random universe generator, and if I'm asked: "do you think you're more interesting than a void?" my answer is that that's worth thinking about but, ultimately, no - it's all just 'stuff' in different arrangements. Yes, I accept that, as the living kind of stuff, I notice a selfish difference, but I don't know how to distinguish this from spooky but meaningless coincidences.
Let me try it out this way. Suppose you were someone who never had dreams. For whatever reason, whenever you sleep it’s always nothingness. Then one night, you have a dream, and the next day your dream comes true. Are you going to attribute special significance to the dream?
I think this is closer to what’s being asked in the FTA, where the non-dream nights are the void universes, but while I think I would be more impressed by this circumstance, I also think I would dismiss it because I have so many other commitments in my worldview that would discourage a special interpretation.
So perhaps the more fundamental thing I’m saying is that I don’t accept that the FTA presents a self-contained argument. I think we’re all making different interpretations of it, as refracted through our other commitments, and mine tell me that nothing significant is being said. Maybe, as you said earlier, we just have to agree to part ways on this one, even though we otherwise agree on nearby cases like the Quran in the CMB.
(I'm all the more surprised about the 61 emails! Sorry that I didn't hold to my promise! Do you remember where the conversation started?)
> I find that the only explanation I need to give is that I have subjective biases that sometimes misfire.
I don't think that is sufficient for me, depending on what you mean.
What I'm about to say is a bit vexed by the impossibility of observing a universe which is not fine-tuned. So, again, lets imagine you're just randomly generating universes in simulation and having a look inside. That way we can focus just on the problem of whether it is interesting.
Suppose you are allowed to look at one universe, and it so happens to have life in it, and you find that interesting. You can chalk this up to a subjective bias and look no further for an explanation if and only if the bias is that you were likely to find any arbitrary universe so interesting and you are unaware of this fact. So, your bias might be that you are overestimating how interesting you find this particular universe as compared to possible alternatives. Maybe you would have been equally fascinated by a featureless void.
But, as long as you really are especially fascinated by this universe, and the probability of being so fascinated is extremely low, then there's an impressive coincidence, and we should be seeking explanations.
> In the analogy with the Quran, having a prior interest would be like saying that we care particularly about the Quran but we don't
I think you're misinterpreting the explanandum here. The explanandum is not why the constants have the particular values they have. As you say, any specific settings of the parameters would have a low co-occurrence probability. This is not what makes it interesting. What is interesting is that the parameters we have conspire to give the unvierse the property of being life-permitting. There are many other configurations that might also achieve this -- varying one constant in one way might be compensated for by varying another constant in another way. The problem is just that the vast majority of possible configurations do not appear to be life-permitting.
Focusing on the specific values of the constants reminds me again of the the pebble parable, just finding an arbitrary pebble and being amazed. No, what makes this interesting is more like finding a pebble of solid gold. It doesn't have to be any specific solid gold pebble, just any pebble with this very surprising property. It doesn't have to be the Quran. Any meaningful text would demand an explanation, but all the more so if it is a text with as much significance as the Quran.
> I don't know how to distinguish this from spooky but meaningless coincidences.
You don't have to. Any sufficiently impressive coincidence should send you looking for an explanation. Sometimes there are none. But sometimes there are.
For example, I just had a heating engineer over to replace a zone valve that had failed. It had failed months ago, and we just got around to fixing it, because we could make do with the other zone valve. All good. Then, that same night, I noticed that the other zone valve wasn't working. That's a coincidence! I still don't know if it's because the connection came loose because of the repair, or if it just failed randomly the same day. Either is possible, but the coincidence makes me suspicious that it had something to do with the repair. That's how this stuff should work. Coincidences should make you seek an explanation, but you're not guaranteed to find one. The more extreme the coincidence, the less you should be prepared to accept that it just is a coincidence.
For coincidences that are only brought to your attention because of how impressive they are, the explanation is straightforward. It's a big world. All kinds of random stuff happens. Sometimes coincidences happen, and you might hear about really impressive ones. So hearing about a really impressive coincidence *because* it is a really impressive coincidence should give you reason to think that it's a big world with all kinds of random stuff happening. Just like with fine-tuning (on my view). If some really impressive coincidence happens to you, personally, then you should consider whether there is a prank or you're dreaming or something.
> Are you going to attribute special significance to the dream?
Yes.
> while I think I would be more impressed by this circumstance, I also think I would dismiss it because I have so many other commitments in my worldview that would discourage a special interpretation.
That's fine. You're not guaranteed to find an explanation. But you should be trying to think of one. Maybe the dream wasn't as like what transpired as you think. Maybe some vague similarlity has grown in your memory, a bit like how people can read all kinds of stuff into the prophecies of Nostradamus. Or maybe it was a huge coincidence. But you should be impressed by the coincidence and you should be looking for explanations, in proportion to how impressive/improbable the coincidence is.
(I think the conversation started on comments on my old blog -- e.g. https://disagreeableme.blogspot.com/2013/12/the-universe-is-made-of-mathematics.html, and somehow moved into email... so in actual fact we had been talking for longer than the email chain. But I don't see your comments on my blog now so I don't know.)
I think it’s helpful to boil down the talk of coincidences to two clearly stated independent things with a low co-occurrence probability. So, in our analogies, we have:
A. (1) an Earth text and (2) that text in the CMB
B. (1) the golden pebble and (2) that pebble on a regular beach
C. (1) I have my first dream and (2) that dream comes true
D. (1) a zone valve breaks and (2) that zone valve’s pair breaks
In each case, we speak casually about having a subjective impression of a coincidence in need of explanation, but that impression is a consequence of appreciating that (1) and (2) are objective facts with apparently non-overlapping causal histories.
Then the FTA comes along and, at least per your description above, it asks us to consider:
E. (1) a life-permitting universe exists and (2) I find that interesting
In this formulation, I’m essentially getting a NaN error because we’ve got an opinion instead of a fact for (2). If we leave it like that, I can say, sure, this is a neat topic for a coffee morning, but let’s not rush into updating our credences about metaphysical propositions like god and the multiverse.
Alternatively, we can try to amend the formulation, and probably the most sensible option is:
F. (1) the universe has variable parameters and (2) its parameters allow for life
For the sake of argument, we’re accepting the low probability of (2), but it doesn’t seem intelligible to speak of (1) having a probability at all, it’s just the context for anything that happens. So, then we’re just talking about (2) on its own and where has the coincidence gone?
Then we can say that it’s fine just to wonder about a single low probability event but to assess its significance we introduce subjective assessment, which is just (E) again and I’m back to NaN.
I agree that the multiverse response to fine-tuning is trash but it absolutely doesn't require something "Godish" to explain. Also I thought you said that explanation was not the right way to think about the issue?
I think it's fine to talk of 'explanation' if you want to make the point more accessibly. But when we get to evaluating disagreement, we should cut that out and move to Bayesian confirmation.
Do you think there's anything to the idea that, once we start reasoning in Bayesian terms, it may arguably be appropriate to treat knowledge of our own existence as background knowledge? I've already massively disconfirmed all versions of naturalism on which I don't exist in the way I appear to (as a biological organism). It seems to me that if I were to wake up in a white room, and I haven't gone and checked the physical parameters yet but I know I'm a biological organism, then I'm going to reason like this: if naturalism is true, I will almost certainly observe life-permitting parameters, but if theism is true, I might not observe life-permitting parameters since God could've supernaturally created me in this room. Then I go and observe life-permitting parameters. In that framing where your knowledge of your own existence is already part of the background, the data seems to support naturalism. Some creationists like to argue that abiogenesis is so implausible that a supernatural creator would need to intervene in the process to produce living cells. So it seems like at least some theists would agree that if we found ourselves in a non-life-permitting universe, that would be evidence for theism and miraculous intervention. But it can't be that finding yourself in a life-permitting universe is evidence for theism and finding yourself in a non-life-permitting universe is also evidence for theism, it would just be that life-permittingness is evidentially neutral towards theism. And then what you're left with is no longer fine-tuning data, it's just the fact that you exist.
If not God, what are your candidates?
Sorry my first reply was horribly worded. I'll try again: Jane's situation doesn't require an explanation. You cannot 'explain' pure statistical probabilities in the same way you can look for explanations for other unusual events that happen in life. When you win the lottery, you don't hand back your winning ticket and refuse to cash it because it clearly couldn't have happened as the odds were so against it. Statistical probabilities are literally not things we can look for scientific explanations for. The are a different category of things, to things such as everyday physical observations.
In the same way the multiverse people are very misguided to think they need to explain life in the universe, the same is true for the fine tuning argument. Multiversers are forced to posit an "infinity" of multiverses in their multiverse theory, because anything less than infinity mutliverses creates glaring logical and statisitical issues in the theory (namely - each individual universe starts with the exact same odds of life, not more or less). But they need'nt bother, there's no required explanation for merely a large statistical probability.
Similarly, fine-tuners think that the parameters of the universe combined with the existence of life "requires" some kind of rational explanation, and I'm not sure why. You can't get a rational scientific explanation of why you won the lottery, or why your sperm beat millions to reach the egg. Your personal phsical existence in time now is already SO unlikely in regards to pure mathematical odds we don't need to appeal to fine tuning of the universe, we've *already reached* numbers that we usually accept as being *physically impossible* just by being born. So that's why we cannot use statistical probabilties in this way. It's a misunderstanding of statistics. Incidentally there's some good modern philosophy out there on this subject of statistics!
Ok and what god created the universe that the god that created this universe existed in?
That must had been a universe funetuned for a god to exist in.
Explanation has to stop somewhere. This might be a good worry if you responding to some kind of 'Why is there something rather than nothing' argument, but it's irrelevant to a Bayesian argument.
Thanks for your answer Professor.
I see your point.
Something completely different.
I see many people who get good replies to their posts but ignore them even if they ought to have the time to reply to them. I'm not talking about someone who have received hundreds of replies. But people who get 5-10 good replies on one of their two posts that day, yet ignore to answer any of them. You seem to be the exact opposite. Really nice to see for a change someone who doesn't come of as elitist.
The crowd-summoning analogy is interesting, but it presupposes a pre-existing pool of possible observers who are assigned to an observed success. Structurally, that is much closer to a Cartesian-dualist or SIA-like setup than to a physicalist one. In my preprint, I argue that in just such a case one indeed gets confirmation of a multiverse over a single-universe hypothesis independently of fine-tuning.
The Jane case, by contrast, is much closer to physicalism. Here, observers are themselves outcomes of universes, not pre-existing candidates for assignment. In that case, if one includes the total evidence while still accounting for selection effects, the calculation shows that there is still Bayesian confirmation of a multiverse over a single-universe hypothesis, but it remains dependent on fine-tuning. See my preprint for the calculation: https://philpapers.org/rec/NABTMV
On the first point, I think you're taking the analogy a bit too literally. It's just a way of capturing the improbability of me in particular existing, as opposed to some other possible creature.
I still haven't had chance to read this, I'm afraid. Maybe check out the maths Kenny Boyce has done for the paper we're working on together (he out the basic maths on his webstie) and see where we part company.
What I’m taking literally is the probabilistic structure: (i) a fixed pool of possible observers, and (ii) any of them can end up in one of the unlikely scenarios. From a probabilistic point of view, that is exactly an SIA-style setup: a fixed pool of possible observers who can end up in any universe. Structurally, that corresponds to a Cartesian-dualist scenario. So it should not be surprising that the crowd-summoning case yields the same result as SIA: confirmation of a multiverse over a single-universe hypothesis independently of fine-tuning. I therefore do not dispute Kenny’s calculation; my point is that this setup already builds in the stronger self-locating assumptions.
The Jane 3 case is different, because there the observer is an outcome of the process, so the pool of observers depends on the available processes. If the evidential claim is that someone with my evidence exists — the issue discussed inconclusively in the video around 1:14, where Kenny says he’d “like to see that build out a bit more” — then, on my calculation, this still yields confirmation of a multiverse over a single universe, but now in a way that remains dependent on fine-tuning.
I would argue that the fine-tuning of our universe gives us evidence for the multiverse (with varying constants) regardless of the fact that the multiverse hypothesis can't explain why our universe, in particular, is fine-tuned.
First, we have to keep in my mind that the surprising fact (the thing that's begging for an explanation) about living in a fine-tuned universe is that the universe is fine-tuned for LIFE and NOT that it's fine-tuned for US specifically. The universe being fine-tuned for life doesn't necessarily entail that it's fine-tuned for us (in the sense of allowing us to potentially exist in it, which also includes the fact that we can only exist in one specific universe).
This already makes the fine-tuning case non-analogous to Jane. What Jane is trying to explain is HER fertilization, not the fact of fertilization itself. It's the latter case that would be analogous to the case of fine-tuning for life, where the thing to be explained is the existence of fine-tuning for LIFE, not US specifically. So, the real question is: by being born and knowing that the odds of fertilization of any given egg are low, does Jane get support for the hypothesis that there must be multiple eggs out there assuming that she's not trying to explain the odds of HER egg getting fertilized but the fact of fertilization itself?
But what about the requirement of total evidence? Shouldn't we include the fact that OUR universe is fine-tuned when evaluating competing hypotheses (multiverse or theism)? Yes, but note that when we include that piece of evidence we're no longer making inferences just from living in a universe fine-tuned for LIFE but also from living in a universe fine-tuned for US. If we're only considering the fact that we live in a universe fine-tuned for life, we can get to the multiverse (the fact that it happens to be our universe that's fine-tuned is neither here nor there), but if we add the fact of living in universe fine-tuned for us, the multiverse seems like a poor explanation.
However, I don't think single-universe theism fares any better than the multiverse! We have to change a couple assumptions in our reasoning to see this though.
If I'm not mistaken, the inverse gambler's fallacy charge against the multiverse hypothesis goes something like this: My total evidence is that I live in a universe fine-tuned for life and, GIVEN THAT MY UNIVERSE EXISTS, a multiverse with randomly varying constants does not give a high probability for MY universe being fine-tuned for life. No matter how many universes there are, the chances of mine being fine-tuned for life are low whereas single-universe theism gives way better odds.
However, the multiverse theorist makes different assumptions. What seems surprising to them about living in a universe fine-tuned for life is the existence of fine-tuning for LIFE, NOT that THEIR universe specifically is fine-tuned. They will then reason like this: If there is a multiverse, the probability of the existence of fine-tuning for life is high and, GIVEN THAT I EXIST, the fact that my universe is fine-tuned is not surprising either (this is the selection effect).
It looks to me like taking MY existence vs MY UNIVERSE'S existence as given in our reasoning leads to radically different conclusions here (in addition to the fine-tuning for life vs us difference). If I take the existence of my universe as given, it seems surprising that it's fine-tuned for life, but if I take my existence as given, that surprise dissipates. The latter supports the multiverse and single-universe theism whereas the former only supports single-universe theism.
If we drop both givens and try to explain why our universe both exists and is fine-tuned for life (therefore also allowing us to exist), both hypotheses, the multiverse and single-universe theism, fare quite badly. The multiverse arguably gives our universe higher chances of existing but fails to explain why it's fine-tuned. Single-universe theism explains why our universe is fine-tuned but not why it exists. It doesn't seem likely that out of the infinite amount of possible fine-tuned universes (with universe-specific observers) that God could have created, he decided to create ours (and us) specifically.
So, perhaps multiverse theism fares the best? I don't know.
To conclude:
1) We CAN get from the existence of fine-tuning for life to the multiverse. It's a logical and legitimate inference to make. However, if we include both pieces of evidence, that fine-tuning for life exists and that our universe is fine-tuned for life, the multiverse is not a great explanation.
2) The inverse gambler's fallacy charge works only if we merge the two concepts (fine-tuning for life vs our universe's fine-tuning) and we take the existence of our universe as given and try to reason from its being fine-tuned to the multiverse. If we keep them separate and take our existence as given, the multiverse hypothesis is solid because there's only one surprising fact to be explained, the existence of fine-tuning for life.
3) If we try to explain why our universe exists AND is fine-tuned for life, the multiverse and single-universe theism are equal(ly bad) explanations.
(And yes, I created a substack account just to reply to this...)
Strange how far some people will go to rationalize a randomness in creation to avoid the possibility of intention.
Jane seems disanalogous to me because the doctor scenario is just bonkers and presumes she is an immaterial spirit that can drop into any egg. Multiverse, on the other hand is a hypothesis that predates fine tuning and for which there are independent reasons to believe.
As someone who grew up around Eastern ideas of karma and purpose, I find your work on cosmic fine-tuning strangely familiar.
The question “Is this all an accident, or is there a telos?” appears both in physics and in people’s private narratives about their lives.
What I appreciate here is your willingness to sit with that question analytically, without collapsing into either dogmatic design or nihilistic randomness.
It's just a question of following the evidence where it leads.
Is there any evidence that the universe could have been fine-tuned in any other way than it currently is?
The probabilities at play are epistemic rather than objective.
Why would something godish choose to create a universe in which the parameters that allow naturalistic life to exist are extraordinarily unlikely?
The God I believe in didn't have a choice about that.
Is that because his limitations mean he can't choose from other options, or other options don't exist? Or both?
Why not?
Sean Carroll makes this kind of argument, but I find it pretty weak. Suppose God had created the universe with other constants, and had just allowed planets to form and the sun to give energy by magic.
However things end up working, they're going to have to work in some logically consistent way. There are going to be regularities. We can do science on it. The net effect would be that the apparent laws of physics would look different than the present laws of physics but with life-incompatible constants. It's going to look suspiciously like everything is fine-tuned to support life. Much like this universe, perhaps.
There are other responses too. Another one would be just to go back to the "why not?". If God can do anything, then all options are equally easy, so God may as well pick from them according to criteria that may as well be aesthetic. Why think that God prefers a magic aesthetic where there's a sort of Rube Goldberg machine set up to allow life to exist despite the basic incompatibility of the rest of the laws of physics as opposed to the elegance of coming up with one set of laws that just conspires to allow life to evolve?
Philip's own response is just that it's a Godish thing of limited powers. The structure of the laws of physics is set somehow. It's metaphysically or logically necessary, perhaps. He doesn't say or know where they come from. But it's just the same problem for him as it is for naturalism, so it's not worse. In Philip's hypothesis, God only has the freedom to set the constants, and if that is the case, there are very few possible tunings that allow life to exist, so those are the values God chooses.
I agree mostly, I think. But then the god that is most popular has created non-physical life, presumably not bound by those restrictions. Are Heaven and Hell logically consistent? Why create a physical realm at all?
Given the statistics of the tuning, asking why not? just prompts me to just say sure... but why tho? If it needs an explanation, it needs an explanation...
Then there's the limited power god, well OK but what makes it a god? All bets are off in that case, why assume he has control of the parameters? They could be fixed by other laws. This being could be an advanced scientist who created our universe entirely by mistake. Or maybe not a mistake, maybe he made a Doomsday weapon to destroy his universe by replacing it with ours. The possibilities could be endless.
May have missed your points entirely, just thinking out loud at this point 🙂
I'd just say that fine-tuning is (arguably) evidence for something Godish, not necessarily for the most popular God. It may not be very strong evidence for any particular account, but it may be evidence against the default assumption that there is just one, naturalistic universe. So, yeah, God could be a scientist who created a universe in a lab, or God could be some teenager in a basement with a supercomputer capable of simulating a universe. Philip has independent reasons for favouring his account over others. Fine-tuning alone won't cut it.
True, and I'm pretty sure I wouldn't be convinced by the other reasons either (moral objectivity I'm assuming).
If I were a god of limited power, who just lights the blue touch paper and stands back, I would want to give my baby the best chance of developing as I want, so in choosing the parameters I would more likely pick a set of options that did so more often than not. Seems safer than what we appear to have. What if I got one of the numbers ever so very slightly wrong?
Well, Philip’s idea is that the structure of the laws are fixed and God(ish) only has the ablility to change the parameters. It’s the nature of the structure of the laws that makes the existence of life so precarious. God(ish) couldn’t have done much to make life more probable than it is, within those constraints.
So Philip is assuming that Mr. God(ish) can only alter the parameters of the physical laws, not the structure... sounds a bit like Philip is fine-tuning God..?
This is the difference between a homogeneous multiverse and a heterogeneous multiverse, right? God(ish) can only choose from the options on the menu of the homogeneous one, for some reason.
Jane is right: her existence is evidence for the existence of the relevant class of observer-producing outcomes, and that class is more likely under a multiverse hypothesis than under a single-universe hypothesis. So a multiverse can be confirmed even if the probability of *her* existence is no higher.